It is suggested that because of the rapid lowering of the viscosity of its solutions by reducing agents the major, important fraction of glutenin consists of linear molecules made up of polypeptide chains linked to one another difunctionally by S.S bonds. When dough is stretched the natural tendency of the polypeptide chains to return to a contracted state of low free energy accounts for the elasticity. The interchain S.S bonds are essential for elasticity because inter-chain adhesion between individual polypeptide chains will not overcome the stronger intra-chain forces unless reinforced by the S.S bond. If extension exceeds the elastic limit, viscous flow occurs because steric hindrance and molecular slip will prevent a return to the original conformation. Disulphide interchange is believed to play an important part in stress relaxation. Mechanical scission of S.S probably occurs when molecules at their elastic limit are subjected to too much stress; this may explain the work maximum in the Chorleywood Bread Process. Explanations are advanced for mechanical and activated dough development and evidence in favour of the hypothesis is discussed. IntroductionGlutenin molecules show a wide variation in size, but their weight-average molecular weight is high, exceeding 1 million according to Jones et aZ.I.2 They are assemblages of polypeptide chains of mol. wt. 20,000, which are joined to one another by S.S bonds from evidence provided by Nielsen et aZ.3 It was concluded by these last workers that the elastic and cohesive properties of glutenin depended on the integrity of the S.S crosslinked structure. The polypeptide chains, referred to throughout this paper as 'chains', though of similar mol. wt., do not appear to be identical because after reductive scission of S.S bonds they show a range of mobilities

The hypothesis that glutenin molecules consist of polypeptide chains joined by SS bonds into linear concatenations, instead of into branching giant molecules, is further examined. The concept that secondary forces can build up sequentially to produce appreciable tension in the molecules appears to fit the facts better than the idea that entangled molecules behave like a knot, which predicts a relation between tenacity and mol.wt. not obseived for other high polymers. Work hardening arises because orientation makes the most effective use of secondary forces, the molecules being aligned so that they overlap by substantial fractions of their lengths. SS interchange in dough not only relieves stress but controls the average length of concatenations. Viscous flow depends predominantly on molecular slip, but is assisted by mechanical fission and SS inteichange. In a resting dough mechanical scission is absent and SS interchange makes a relatively greater contribution to stress relaxation. Mechanical scission will not occur in terminal segments of concatenations. In an overworked dough, the length of many concatenations is at least halved by mechanical scission, greatly reducing the resistance. The predicted level of SH groups produced by mechanical fission is of a similar order to an experimental value quoted in the literature.

The fall in viscosity of dispersions of gluten when treated with excess of mercaptoethanol does not exhibit an initial induction period. This finding appears to rule out the idea that glutenin is highly crosslinked in a branching mode. It is compatible with a linear model in which one SS bond, not two, joins adjacent chains, or a model with low levels of branching. However, it is unnecessary to postulate branching crosslinks to explain glutenin properties in the present state of knowledge. The insoluble residue fraction of wheat flour protein is believed to be mainly linear glutenin of high molecular weight. A linear hypothesis can account for the Orth and Bushuk effect, in which baking quality is positively and negatively correiated with insoluble and soluble protein, respectively. It is suggested that gliadin may act as a plasticiser and aid to dispersion of glutenin.

Cereal storage proteins, it is suggested, are more tolerant of genetic mutations than other proteins. Evolutionary tendencies may have been towards decreasing the cystine content and increasing the content of amino acids which are easily synthesised, especially those which are rich in nitrogen. The appearance of stress in molecular structure so that an intrachain SS bond is no longer able to close, thus creating interchain linkages, may have led to the origin of the glutelins.It is argued that a viscoelastic dough can be formed only if protein particles, which originate in the form of protein bodies, are able to form a continuous structure whether maintained by covalent bonds orland secondary forces. The constituent polypeptide chains of glutelins could be joined in a linear or a branched mode. A number of difficulties arise if the branched model is used to explain the properties of dough at the molecular level. In attempting to reconcile fact and theory, it is speculated that polypeptide chains of wheat, rye and barley glutelins form linear concatenations with two disulphide bonds connecting each chain to the next: continuity of structure in their doughs depends on entanglement and interaction by secondary forces. The glutelin molecules of oats and maize are unable to form a continuous network, hence the absence of viscoelasticity from doughs of these cereals. Two reasons are suggested: more probably the molecules are highly cross-linked in the branched mode and are thus incapable of significant entanglement; or if the molecules were cross-linked in the linear mode, poor solubility would ensure that protein-protein contacts in protein bodies were maintained, thus preventing dispersal of the protein and subsequent molecular entanglement. The glutelin molecules of sorghum appear to be too small to create a network by molecular entanglement. The importance of the disulphide interchange reaction is thought to lie in stress relaxation, but not in the creation of a dough structure. It is suggested that SS groups in viscoelastic doughs are sterically protected except when exposed by stress and consequently stress relaxation by SS interchange tends to occur only when it is needed.

Amino acid analyses have been made on flours of wheat, rye, barley, oats and maize. The overall pattern is sufficiently similar (particularly between rye and barley) to indicate a common ancestor, but significant individual differences occur. Wheat protein differs from the other four in its higher capacity for polar and H‐bonds and lower content of salt links. Other factors such as S.S interchange potential may be more important than the pattern of intermolecular forces since the latter does not change much from rye to barley in spite of their rheological differences. The factors for conversion of nitrogen to protein are, in the above order, 5.7, 5.8, 5.8, 5.7, 5.9.